Creating annotation for visuals [Technical Preview]

Annotations are brief summaries or description that you can associate with a visual or dashboard. You can use them to clarify data details, highlight outlier values, or provide additional context for visual consumers. The annotations feature (currently in Technical Preview) supports both manually written descriptions and AI-generated summaries based on the data displayed in a visual.

How annotations appear🔗

If a visual has no annotation:

In View mode, the Annotation icon () is hidden.
In Edit mode, the Annotation icon is visible. Clicking it displays a message indicating that no annotations have been added yet, and provides the Edit option for adding an annotation.

If a visual has an annotation:

In View mode, the Annotation icon is visible. Clicking it displays the annotation.
In Edit mode, the Annotation icon is visible. Clicking it displays the annotation, and provides the Edit option for editing the annotation.

You can create or edit annotations only in Edit mode.
To use AI-powered summarization, enable AI features and AI summary in Site Settings. If these are not enabled, only manual annotations can be created. For more information on these settings, see Managing AI settings.

Creating an AI-powered annotation🔗

With AI-powered summarization, the system analyzes the visual and automatically generates a concise summary for you. The AI annotation leverages generative AI to simplify the interpretation of complex visuals.

Open a dashboard in Edit mode.
Click in the top-right corner of the visual you want to annotate and click Edit.

The Annotations modal window is displayed.
Enable the Generate with AI toggle.

Information about the AI prompt is displayed.
Click View Prompt to review or edit the default AI prompt text.

You can also restore the default prompt if needed.
Click Generate with AI.

The annotation is automatically generated and a Generated with AI label is added.
important
If the dataset contains sensitive fields, a warning appears to inform you that sensitive fields are not sent to the AI, which may affect the annotation’s accuracy.
Review the generated annotation for accuracy.
note
AI-generated content may be inaccurate or misleading. Always verify the content.
[Optional] Edit the annotation text.

If you edit the text, the Generated with AI label is removed.
Click Save.

Creating a manual annotation🔗

Open a dashboard in Edit mode.
Click in the top-right corner of the visual that you want to annotate and click Edit.

The Annotations modal window is displayed.
Enter your annotation text in the text box.
Click Save.